Struggling with spatial omics analysis? Try Sopa, our new technology-invariant Python library for all image-based spatial-omics! Based on @scverse_team data structures, and built with @KevinMulder19 and @FGinhoux 🚀 https://t.co/e9XA7vHvY4

Happy to share Novae, our new foundation model for spatial transcriptomics data 💫 Load a pre-trained model and start using Novae in two lines of code! https://t.co/4r2kEWhYFU https://t.co/q8NChl5Jau

Sopa is now published, read the paper and try it out on your own spatial data!👇 https://t.co/mCXxedikAE

6/6 Check out Sopa Our preprint: https://t.co/6pRaWIqlqS Our documentation:

5/6 Spatial analysis and niches geometries Sopa allows a fast computation of many geometric characteristics related to the niches (or spatial domains). It also computes distances from cell-types to cell-types, cell-type to niches, or niches to niches

4/6 Example on multiplex imaging data Tested on a MACSima and a PhenoCycler dataset, Sopa shows a high resolution (each cell is represented by the mean intensity of all stainings inside their boundaries)

3/6 Example on spatial-transcriptomics data Tested on a MERSCOPE and a Xenium dataset, Sopa shows a higher resolution compared to the proprietary segmentations

2/6 Memory efficiency Sopa can run on large datasets using less than 16GB of RAM. The time-consuming and RAM-consuming operations are optimized

1/6 Overview of the pipeline Sopa uses the SpatialData data structure, and performs segmentation, aggregation, and annotation. The pipeline outputs contain (i) Xenium Explorer files for interactive visualization, (ii) a QC report, and (iii) a data directory for further analyses